The following are the additions that will be done by Visual Studio for us when creating a web or worker role in Azure.

1. WebRole.cs

2. Web.Config

We have to do the following.

1. Add an Azure Storage account to the Service Configuration file.

            Where the ACCOUNT_NAME is the name of the blob service and ACCOUNT_KEY is the primary or secondary access key.

2. Update the Diagnostic Monitor configuration to update configuration settings like the transfer period schedule from the machine to blob or the log level filter. This has to be done in the WebRole.cs before calling the Start method within OnStart.

We collect the default configuration and update the settings to start the Diagnostic Monitor with the new settings.  Logs are the internal trace logs that will be saved in the WADLogsTable within the blob tables. In order to collect the event log, we will add  an event log data source to it as well.

This will collect Windows Event Log entries that are within the Application channel and belong to "System Restore". The Windows Event Logs will be saved in WADWindowsEventLogsTable  within the blob tables.

Additional Details on configuring the Diagnostic Monitor can be found here

3. The last task is to add the code to actually trace or log the messages. The .net base classes under System.Diagnostics are supported for Windows Azure. You can use Trace class for the Logs and EventLog class for the Windows Event Logs.

Note: If you want to write an entry into the event log, you have to use an existing source. Creating a new source requires admin rights which are not supported in Azure.

While migrating my stored procedures to SQL Azure, I received errors in some because of lack of support of some of the functionality. I had to find a workaround to successfully export them to SQL Azure.

Working with XML

sp_xml_preparedocument

It reads the XML string within the SP, parses it using MSXML parser and provides a handle to the parsed document, which is stored locally by the SQL Server. This parsed document is a tree representation of the various nodes in the XML document. It is not supported by SQL Azure.

sp_xml_removedocument

It removes the parsed xml document created as a result of sp_xml_preparedocument. It is not supported by SQL Azure.

OpenXML

It provides a rowset view over an xml document. It is not supported by SQL Azure.

Example

Output

Workaround

The workaround is to use nodes(..) and value(..) functions

As you can see in the highlighted portion, all the details of the WITH in the previous example have been passed within value(..) function, which is used in the SELECT statement. Also, the sp_xml_preparedocument and sp_removedocument have been removed. This method directly uses the xmlString to get the rows.

Now, press F5 and you will get the following.

Don't close the existing solution. Copy the URL address and open a new Visual Studio environment with a new solution and a Console project. Add a service reference to it, providing the above URL.

FactorialReference is the namespace given to this web service reference. Use it to calculate the factorial and output the result. Press F5 and you will see the command window output "The factorial of 5 is 120".

Why did we do that, you will know in the end. Now, stop both the solutions and publish the web service just the way mentioned in the Windows Azure: Hello World Part 2.

Once you have successfully deployed and run the web service, it is time to update the reference of the console application. Open the "app.config" of the console project and modify the client endpoint address to set to the production URL of the web service in cloud.

We did not directly update the service reference in the console project to point to cloud directly because that does not work. This is because of some configuration of internal Windows Azure environment to handle internal and external comms pathways that needs to be fixed. But since we have a workaround, we don’t need to worry about that.

We start by establishing a connection from our SQL Server Management Studio to the SQL Azure. But for that, we must first allow our computer to be connected to SQL Azure, since SQL Azure restricts all communication as a default setting. For that, Login to your SQL Azure account.

Go to Firewall Settings and check "Allow Microsoft Services access to this server. This will allow communication from Windows Azure services.

Next we want to add a rule to give access to our local machine. Click on "Add Rule" and the popup window will show up having the ip address of the local machine. Since I have 2 machines, one is 103 and one is 117, so I will give a wider limit. Once you are done, click Submit.

Now if you go to the "Databases" tab and select the "master" database, you can test its connectivity as well as view its connection string. I would recommend you to copy the connection string into your notepad for future reference.

Open your SQL Server 2008 R2 Management Studio and enter your SQL Azure database name (the complete name) into the Server name box and select SQL Server Authentication in Authentication. Then provide the SQL Azure Administrator username and password and click "Connect". Remember that the "Server type" has to be "Database Engine".

Note: For SQL Server 2008, you have to close the connection window and click "New Query" to connect successfully.

Once you are connected, you will not see any databases except for the master.

Now Launch your SQL Azure Data Sync tool to export the database schema from local to the SQL Azure database.

Enter the required SQL Azure Server info and test the connection. If it is succeeded, move on.

Enter the local SQL Server info and test the connection. If it is succeeded, move on.

Select the tables that you want to export. Since we are exporting the database to SQL Azure for the first time, I would set the "Conflict Resolution Policy" to Local. It means that in case of conflict, the final data would be of the local database.

Rearrange the order of the tables, with the independent tables at the top, such that all the tables must have their dependent tables above them.

Review the changes and click Process to start synchronizing the two databases.

Once the synchronization is complete, go to the SQL Server Management Studio and you will see an EmployeeVacation database in SQL Azure.

The SQL Azure data sync tool created a SQL Server Agent sync job that runs once a day and synchronizes both your databases. This job is created on the SQL Server Agent of your local SQL Server containing the local database. However, in case of conflict, the data is taken from the local server. If your cloud data is your production data then you should set Azure in "Conflict Resolution Policy" in the 3rd Step of SQL Azure data sync tool.

If you want to run this sync now, you can right click the job and select "Start job at Step…" and it will sync your data.

Note: At present, SQL Azure Sync Framework and Tools installation has to be 32-bit (x86), even on 64-bit machines, since x64 version creates problems while synchronizing and your job will most likely fail.

Check the data in SQL Azure database and you will see that it is synchronized with the local database.

Right click "HelloWorldService" project and click "Publish". It will open a folder with two files in it, a package file with the extension .cspkg and a configuration file with the extension .cscfg.

Sign into your windows azure account by going online at https://windows.azure.com

Create a new service by clicking "New Service". This service should be a hosted service, since we want to host an ASP.NET application.

Provide a name and description for it and click "Next".

Enter a public name for your service. Also since we don’t have any dependency with other applications, it doesn’t matter where it is stored.

Our service is now ready to receive any deployment. We first try to deploy our application in the Staging area just to test whether it works or not. Once it is successfully deployed over the Staging area, we can easily move it to the Production.

In the Staging Deployment, upload the .cspkg file published by Visual Studio in the "Application Package" and the .cscfg file in the "Configuration Settings". Also, give this deployment a name.

Once you are done, click "Deploy" and it will start deploying the application.

The application will be stopped initially after deployment. You have to click "Run" to make it online.

The status will be "Initializing" in the beginning. It will take some time for the application to be online.

Once the status is "Ready", you can connect to your website using the "Web Site URL" link. You can update and re-configure your website if you need some tweaks here and there. Once you are satisfied with it, you can move it over to Production by pressing the blue button in the center.

Congratulations, you have just created your own very first Hello World application in cloud.

Open Visual Studio in administrator mode (Run as administrator) and create a new "Windows Azure Cloud Service" project.

Add an ASP.NET web role to the project.

In the default.aspx file, type in some text  within the "div" tags.

If you press F5 now, your application should now run. If it gives you the following error,

It means that you don’t have your development storage setup. For now, just turn it off from the following

HelloWorldService project properties --> Development

The output will give you the following.

This project is run by imitating the windows azure deployment in local computer. You can look at the output by going to the Development Fabric window.

Now we are ready to deploy our Hello World application on the cloud here.